 Alexander Borzunov
|
93c4eba5d1
Bump version to 1.1.4 (#306)
|
2 jaren geleden |
|
Alexander Borzunov
|
c0e0e1319d
Force transformers to use config.torch_dtype by default (#307)
|
2 jaren geleden |
|
Alexander Borzunov
|
98be9ffe4c
Relax the rest of Hugging Face dependencies (#305)
|
2 jaren geleden |
|
Alexander Borzunov
|
5c0b4286b2
Suggest commands for Docker first (#304)
|
2 jaren geleden |
|
Alexander Borzunov
|
35662b4a16
Require bitsandbytes == 0.38.0.post2, hivemind == 1.1.7 (#302)
|
2 jaren geleden |
|
Alexander Borzunov
|
21c3526ec1
Start SequenceManager's thread only after first .make_sequence() (#301)
|
2 jaren geleden |
|
Alexander Borzunov
|
6c6150f684
Remove use_auto_relay=True in client (#300)
|
2 jaren geleden |
|
Alexander Borzunov
|
892fa2386a
Remove CustomLinear8bitLt (#297)
|
2 jaren geleden |
|
Alexander Borzunov
|
74d8cda8c4
Add Python 3.10 to CI (#299)
|
2 jaren geleden |
|
Alexander Borzunov
|
2116df08bc
Fix deps, enable 8-bit by default for TP (#298)
|
2 jaren geleden |
 justheuristic
|
987f4d2b2f
Update bitsandbytes, hivemind, transformers (#290)
|
2 jaren geleden |
|
Alexander Borzunov
|
e0cef73757
Hotfix: Increase daemon_startup_timeout (#292)
|
2 jaren geleden |
|
Alexander Borzunov
|
a7d3d02194
Fix invalid author email in setup.cfg (#287)
|
2 jaren geleden |
|
Alexander Borzunov
|
8dab37c1a9
Add benchmarks to readme (#284)
|
2 jaren geleden |
 Max Ryabinin
|
793726b041
Speed up loading blocks using init with meta weights (#285)
|
2 jaren geleden |
|
Alexander Borzunov
|
c519bffc59
Bump version to 1.1.3 (#278)
|
2 jaren geleden |
|
Alexander Borzunov
|
aae1f4f368
Increase default request_timeout (#276)
|
2 jaren geleden |
|
justheuristic
|
fb2583b682
Use inference mode in _MergedInferenceStep (#275)
|
2 jaren geleden |
|
Alexander Borzunov
|
fd9400b392
Fix use_chunked_forward="auto" on non-x86_64 machines (#267)
|
2 jaren geleden |
|
Alexander Borzunov
|
a2e7f27a5a
Improve "connect your GPU" message (#266)
|
2 jaren geleden |
|
Alexander Borzunov
|
fee19e9b9b
Use get_logger(__name__) instead of get_logger(__file__) (#265)
|
2 jaren geleden |
|
Alexander Borzunov
|
55e7dc07a0
Limit max delay between retries to 15 min (#264)
|
2 jaren geleden |
|
Alexander Borzunov
|
38b071135b
Show visible maddrs for public swarm too (#263)
|
2 jaren geleden |
|
Alexander Borzunov
|
42594e5173
Link FAQ in readme (#260)
|
2 jaren geleden |
|
Alexander Borzunov
|
2a5070aa1a
Improve reachability logs (#253)
|
2 jaren geleden |
|
Alexander Borzunov
|
4091db10bf
Lower payload size threshold for stream handlers (#251)
|
2 jaren geleden |
|
Alexander Borzunov
|
9954cb84fe
Add `allowed_servers`, `max_retries` options to the client, improve logs (#235)
|
2 jaren geleden |
|
Alexander Borzunov
|
3c523ab0d2
Fix TP crashing when hypo_ids are used (#249)
|
2 jaren geleden |
|
justheuristic
|
b8a6788490
Fix examples/sst, add cls_model embeddings (#248)
|
2 jaren geleden |
|
justheuristic
|
8766a14d28
Minor changes to examples/prompt-tuning notebooks (#247)
|
2 jaren geleden |
